ConsInstancy: Learning Instance Representations for Semi-Supervised Panoptic Segmentation of Concrete Aggregate Particles

TL;DR

This paper introduces ConsInstancy, a semi-supervised panoptic segmentation method that uses novel instance representations and consistency regularization to improve segmentation accuracy on concrete particle datasets, leveraging unlabeled data effectively.

Contribution

The paper proposes a new semi-supervised learning framework with novel instance representations and consistency regularization for panoptic segmentation, specifically applied to concrete aggregate particles.

Findings

Achieves up to 5% higher overall accuracy with unlabeled data.

Outperforms state-of-the-art semi-supervised segmentation methods.

Demonstrates effectiveness on concrete datasets, including a new dataset for fresh concrete.

Abstract

We present a semi-supervised method for panoptic segmentation based on ConsInstancy regularisation, a novel strategy for semi-supervised learning. It leverages completely unlabelled data by enforcing consistency between predicted instance representations and semantic segmentations during training in order to improve the segmentation performance. To this end, we also propose new types of instance representations that can be predicted by one simple forward path through a fully convolutional network (FCN), delivering a convenient and simple-to-train framework for panoptic segmentation. More specifically, we propose the prediction of a three-dimensional instance orientation map as intermediate representation and two complementary distance transform maps as final representation, providing unique instance representations for a panoptic segmentation. We test our method on two challenging data sets of both, hardened and fresh concrete, the latter being proposed by the authors in this paper demonstrating the effectiveness of our approach, outperforming the results achieved by state-of-the-art methods for semi-supervised segmentation. In particular, we are able to show that by leveraging completely unlabeled data in our semi-supervised approach the achieved overall accuracy (OA) is increased by up to 5% compared to an entirely supervised training using only labeled data. Furthermore, we exceed the OA achieved by state-of-the-art semi-supervised methods by up to 1.5%.